Desktop version

Home arrow Economics

  • Increase font
  • Decrease font

<<   CONTENTS   >>

Concluding Remarks

The joint modeling framework facilitates an integration of multi-source data in the early drug development phase, particularly the associations between chemical structures, biological activity, and gene expressions, in order to identify potential leads in the early phase of drug discovery alongside the development of genomic biomarkers for efficacy of compounds. Selecting and evaluating biomarkers in early drug discovery can substantially shorten development time or the time to reach a critical decision point, such as candidate selection, in drug development.

FIGURE 16.17

The Integrated JM Shiny App. Left Panel: Specification of the variables for the analysis. Right Panel: Top K genes and adjusted/unadjusted association.

The joint modeling approach, although implemented using only one feature at a time for every data source, facilitates the extraction of valuable insights into the associations between chemical structures and mechanism of actions. Although we focused in this chapter on one fingerprint feature and on-target assay per project, this method can easily be run in loops. In a pharmaceutical pipeline implementation, this model can be applied to all or a defined set of interesting chemical substructures, genes, and biological assays (efficacy or toxicity related). The large amount of output can then be collated and filtered for vital information that can help the research team, especially, the medicinal chemist and biologist in taking the next step.

FIGURE 16.18

The Integrated Shiny App: Classification of genes and volcano plot.

<<   CONTENTS   >>

Related topics